Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A system comprising: a host computer system; a backend storage array having a first controller and a second controller; an optical circuit switch (OCS) connected between the host computer system and the first controller and the second controller; a virtualization node system coupled between the host computer system and the OCS to connect the host computer system to the OCS, wherein the virtualization node system includes a first virtualization node and a second virtualization node, each including direct attachment connections to provide clustering of a plurality of servers of the host computer system; and a storage system controller comprising: a failover detector to detect a failover of the first controller when the first controller is in an active state and the second controller is in a passive state; and an OCS controller to control the OCS to switch connection of the host computer system from the first controller to the second controller based on the failover detector detecting a failover of the first controller to place the second controller in an active state.
This invention relates to high-availability storage systems, specifically addressing the challenge of seamless failover in storage arrays to maintain uninterrupted access to data. The system includes a host computer system, a backend storage array with two controllers (a primary active controller and a secondary passive controller), and an optical circuit switch (OCS) that dynamically routes connections between the host and the controllers. A virtualization node system, consisting of two nodes, provides direct attachment clustering for multiple servers in the host system, ensuring efficient load distribution and redundancy. The storage system controller monitors the controllers for failures. If the active controller fails, the failover detector triggers the OCS controller to reroute the host's connection from the failed controller to the secondary controller, which then transitions to an active state. This automated failover mechanism minimizes downtime and data access disruptions, enhancing system reliability. The OCS enables rapid, hardware-level switching, while the virtualization nodes ensure that clustered servers remain connected during the transition. The system is designed for environments requiring continuous availability, such as enterprise data centers or cloud infrastructure.
2. The system of claim 1 , wherein each server is coupled to the OCS through the virtualization node system.
A system for managing server resources in a virtualized computing environment addresses the challenge of efficiently allocating and monitoring server resources across multiple servers. The system includes a plurality of servers, each configured to host virtual machines (VMs) and communicate with an Operations Control System (OCS). The OCS is responsible for managing and monitoring the servers, including tracking resource usage, performance metrics, and operational status. Each server is connected to the OCS through a virtualization node system, which acts as an intermediary layer. This node system facilitates communication between the servers and the OCS, ensuring seamless data exchange and coordination. The virtualization node system may include components such as hypervisors, virtual switches, or other virtualization software that enable the servers to operate in a virtualized environment. By integrating the servers with the OCS through this node system, the system ensures centralized control, improved resource allocation, and enhanced monitoring capabilities. This setup allows for dynamic adjustments to server configurations, load balancing, and efficient utilization of computing resources. The system is particularly useful in data centers or cloud computing environments where multiple servers must be managed efficiently to support virtualized workloads.
3. The system of claim 1 , wherein the direct attachment connections include connections on the first virtualization node to connect the first virtualization node to each of the plurality of servers and to the second virtualization node, and connections on the second virtualization node to connect the second virtualization node to each of the plurality of servers and to the first virtualization node.
A system for managing virtualized computing environments includes multiple servers and at least two virtualization nodes. The system establishes direct attachment connections between the virtualization nodes and the servers, ensuring high-speed, low-latency communication. Each virtualization node is directly connected to every server in the system, as well as to the other virtualization node. This configuration enables efficient data transfer and resource sharing across the network. The direct connections eliminate the need for intermediate routing devices, reducing latency and improving performance. The system is designed to support scalable and resilient virtualized environments, where virtual machines or containers can be dynamically allocated and managed across the servers. The direct attachment architecture ensures that all nodes have equal access to shared resources, enhancing reliability and fault tolerance. This approach is particularly useful in data centers and cloud computing environments where low-latency, high-bandwidth communication is critical for performance-intensive applications.
4. The system of claim 3 , wherein the OCS includes a first OCS and a second OCS, wherein the first OCS is coupled to a first port of the first virtualization network and to the first and second controllers, and the second OCS is coupled to a first port of the second virtualization node and to the first and second controllers.
A system for managing virtualization networks includes multiple Open Compute Servers (OCS) interconnected with virtualization nodes and controllers. The system addresses the challenge of efficiently distributing and managing network traffic in virtualized environments. The OCS units serve as intermediary devices that facilitate communication between virtualization nodes and controllers, ensuring seamless data flow and resource allocation. The system comprises a first OCS and a second OCS, each connected to a distinct virtualization node and both controllers. The first OCS is linked to a first port of the first virtualization node and to both the first and second controllers, while the second OCS is similarly connected to a first port of the second virtualization node and to the same controllers. This dual-OCS configuration enhances redundancy, load balancing, and fault tolerance, ensuring continuous operation even if one OCS or connection fails. The system optimizes network performance by distributing traffic across multiple paths, reducing bottlenecks, and improving overall efficiency in virtualized computing environments. The controllers oversee the OCS units, managing configurations, monitoring performance, and coordinating data transmission to maintain system stability and reliability.
5. The system of claim 4 , wherein the first virtualization node and the second virtualization node each include an additional port to connect to at least one of the group consisting of: an additional back end storage array; and an additional server.
This invention relates to a distributed virtualization system for managing data storage and processing across multiple nodes. The system addresses the challenge of efficiently scaling and interconnecting virtualization nodes to handle increased storage and computational demands without compromising performance or reliability. The system includes at least two virtualization nodes, each capable of interfacing with back-end storage arrays and servers. Each node is equipped with multiple ports to facilitate high-speed data transfer and communication. The nodes are interconnected to form a distributed architecture, enabling load balancing, redundancy, and seamless data access across the network. The system dynamically allocates resources based on demand, optimizing performance and ensuring high availability. Additionally, each virtualization node includes an extra port designed to connect to either an additional back-end storage array or an additional server. This expandability allows the system to scale horizontally by integrating more storage or computational resources as needed. The extra port enhances flexibility, enabling the system to adapt to varying workloads and storage requirements without requiring significant reconfiguration. The system ensures data consistency and integrity through synchronization mechanisms between nodes, preventing conflicts and ensuring reliable access. The architecture supports both storage and compute-intensive applications, making it suitable for enterprise environments requiring robust, scalable virtualization solutions.
6. The system of claim 1 , wherein the OCS comprises an input coupled to an I/O port of the virtualization node system, a first output coupled to the first controller and a second output coupled to the second controller, and a mirror structure to change connections of the input of the OCS from connecting to the first output to connecting to the second output, based on an instruction from the OCS controller.
This invention relates to a system for managing connections in a virtualization node system, particularly for failover or load balancing between redundant controllers. The system includes an Output Connection Switch (OCS) that dynamically routes input data from an I/O port of the virtualization node to one of two redundant controllers. The OCS has an input connected to the I/O port, a first output connected to a first controller, and a second output connected to a second controller. The OCS also includes a mirror structure that can switch the input connection from the first output to the second output, or vice versa, based on an instruction from an OCS controller. This allows seamless redirection of data traffic between controllers, ensuring high availability and fault tolerance in the system. The OCS controller may issue the instruction in response to a detected failure, a load-balancing requirement, or a maintenance operation. The system ensures continuous operation by maintaining redundant paths and enabling rapid failover without interrupting data flow. This approach is particularly useful in high-availability computing environments where uninterrupted service is critical.
7. The system of claim 6 , wherein the mirror structure includes a fixed mirror and a movable mirror structure to operate in response to an instruction from the OCS controller to change the direction of an optical signal received at the input of the OCS from the first output of the OCS to the second output of the OCS.
This invention relates to optical circuit switching (OCS) systems, specifically addressing the need for dynamic redirection of optical signals within a network. The system includes a mirror structure that enables precise control over the path of an optical signal. The mirror structure comprises a fixed mirror and a movable mirror assembly. The movable mirror can adjust its position in response to instructions from an OCS controller, allowing the system to redirect an optical signal from a first output port to a second output port. This functionality enhances flexibility in optical networking by enabling on-demand signal routing without manual intervention. The OCS controller issues commands to the movable mirror, which then alters the signal path accordingly. The fixed mirror provides a stable reference point, ensuring accurate alignment and minimizing signal loss during redirection. This design improves efficiency in optical networks by automating signal routing, reducing latency, and enhancing scalability. The system is particularly useful in high-speed data transmission environments where rapid reconfiguration of optical paths is required.
8. The system of claim 1 , wherein the storage system controller further comprises a virtualization manager to control operations of the first and second virtualization nodes to control clustering of the plurality of servers.
A system for managing and virtualizing server resources in a clustered computing environment addresses the challenge of efficiently coordinating multiple servers to operate as a unified, high-performance computing resource. The system includes a storage system controller that interfaces with a plurality of servers, each running virtualization nodes to abstract and manage physical hardware resources. The storage system controller further comprises a virtualization manager that oversees the operations of these virtualization nodes, ensuring seamless clustering of the servers. This clustering enables the servers to function as a cohesive unit, improving resource utilization, scalability, and fault tolerance. The virtualization manager dynamically allocates and reallocates resources across the clustered servers based on workload demands, optimizing performance and reducing downtime. The system also supports load balancing, failover mechanisms, and centralized management of virtualized resources, enhancing operational efficiency and reliability in data center environments. By integrating virtualization management with storage system control, the system simplifies the deployment and maintenance of large-scale, distributed computing infrastructures.
9. The system of claim 1 , wherein the storage system controller further comprises an OCS direction updater and polling apparatus to verify that I/O signals from the host computer system have been switched to the second controller after the failover detector detects a failover of the first controller.
A storage system controller includes a failover detector that identifies when a primary controller fails and triggers a failover to a secondary controller. The system ensures seamless data access by switching input/output (I/O) signals from a host computer to the secondary controller. To confirm the switch, the controller includes an OCS (Out-of-Service) direction updater and polling apparatus. This component verifies that the I/O signals have been successfully redirected to the secondary controller after failover detection. The polling apparatus periodically checks the status of the I/O signal routing, while the OCS direction updater adjusts the signal paths as needed. This ensures continuous data availability and prevents disruptions during controller failover events. The system is designed for high-availability storage environments where uninterrupted access to data is critical. The failover detection and I/O signal verification processes are automated, reducing manual intervention and minimizing downtime. The OCS direction updater and polling apparatus work together to maintain data integrity and system reliability during controller transitions.
10. A method comprising: detecting, via a backend path detector, a failover of a first controller of a backend storage array connected to a host computer system, when the first controller is in an active state; and performing a switching operation, via an optical circuit switch (OCS), based upon the detecting of the failover of the of the first controller, to connect a second controller of the backend storage array to change the second controller from a passive state to an active state wherein the host computer system comprises a plurality of servers, each connected to the OCS via a virtualization node system coupled between the host computer system and the OCS.
This invention relates to high-availability storage systems, specifically addressing failover mechanisms in backend storage arrays connected to host computer systems. The problem solved is ensuring seamless continuity of storage access when a primary controller in the storage array fails, minimizing downtime and data access disruptions. The system includes a backend storage array with at least two controllers—a first (active) controller and a second (passive) controller. A backend path detector monitors the active controller for failures. Upon detecting a failover of the first controller, an optical circuit switch (OCS) automatically performs a switching operation to disconnect the failed controller and connect the second controller, transitioning it from a passive to an active state. This ensures the host computer system, which comprises multiple servers, maintains uninterrupted access to the storage array. The host computer system is connected to the OCS through a virtualization node system, which acts as an intermediary between the servers and the OCS. This architecture allows the OCS to dynamically reroute connections without requiring manual intervention, improving system reliability and reducing recovery time during controller failures. The solution is particularly useful in enterprise environments where continuous data availability is critical.
11. The method of claim 10 , further comprising determining whether an I/O flow between the host computer system and the first controller continues after failover of the first controller has been detected, and, based on determining that the I/O flow to the first controller continues, initiating a command to the OCS to change direction of the I/O flow to the second controller.
This invention relates to data storage systems, specifically methods for managing input/output (I/O) flow during controller failover in a redundant storage environment. The problem addressed is ensuring seamless data access when a primary storage controller fails, preventing disruptions in I/O operations. The method involves monitoring I/O flows between a host computer system and a primary (first) storage controller. Upon detecting a failover event where the primary controller becomes unavailable, the system checks whether the I/O flow to the primary controller persists despite the failure. If the I/O flow continues, the system initiates a command to an Output Control Switch (OCS) to redirect the I/O flow to a secondary (second) controller. This ensures uninterrupted data access by dynamically rerouting traffic to the backup controller when the primary fails. The method operates within a redundant storage architecture where multiple controllers manage data access to shared storage resources. The OCS acts as a switching mechanism to direct I/O traffic between controllers and the host system. By detecting and responding to persistent I/O flows after a failover, the system prevents data access interruptions and maintains system reliability. This approach is particularly useful in high-availability storage environments where continuous data access is critical.
12. The method of claim 10 , further comprising indicating a cluster of servers in the host computer system based upon confirming that the I/O flow has changed to the second controller.
A system and method for managing input/output (I/O) operations in a host computer system with multiple controllers. The invention addresses the challenge of efficiently distributing I/O workloads across redundant controllers to ensure high availability and performance. The method involves monitoring I/O flows between the host system and storage devices, detecting changes in the I/O flow direction, and dynamically adjusting the cluster of servers handling the I/O operations. When an I/O flow is confirmed to have shifted from a first controller to a second controller, the system identifies and indicates a specific cluster of servers within the host computer system that should handle the I/O operations. This ensures that the I/O workload is properly distributed and that the system maintains optimal performance and reliability. The method may also include steps for initializing I/O flows, detecting flow changes, and validating the new flow direction before adjusting server clusters. The invention is particularly useful in high-availability environments where seamless failover and load balancing are critical.
13. The method of claim 10 , wherein the virtualization node system includes a first virtualization node and a second virtualization node, each including direct attachment connections to provide clustering of the plurality of servers.
This invention relates to virtualization node systems for clustering multiple servers. The problem addressed is the need for efficient and scalable server clustering in virtualized environments, particularly where direct attachment connections between nodes are required to enable high-performance communication and coordination. The system includes at least two virtualization nodes, each capable of direct attachment connections to multiple servers. These direct connections facilitate clustering by enabling low-latency, high-bandwidth communication between the servers and the virtualization nodes. The clustering allows for load balancing, failover support, and resource pooling across the servers. Each virtualization node may manage a subset of the servers, with the direct attachments ensuring seamless integration and coordination between nodes. The system may also include mechanisms for dynamic resource allocation, fault tolerance, and centralized management of the clustered servers. The direct attachment connections may be implemented using high-speed interfaces such as InfiniBand, Ethernet, or other low-latency networking technologies. The clustering configuration ensures that the servers operate as a unified, highly available computing resource, improving performance and reliability in data center or cloud computing environments.
14. The method of claim 13 , wherein the direct attachment connections include connections on the first virtualization node to connect the first virtualization node to each of the plurality of servers and to the second virtualization node, and connections on the second virtualization node to connect the second virtualization node to each of the plurality of servers and to the first virtualization node.
This invention relates to a distributed virtualization system where multiple virtualization nodes are directly connected to a plurality of servers and to each other. The system is designed to improve resource management and communication efficiency in virtualized environments. The problem addressed is the latency and complexity introduced by traditional network-based connections between virtualization nodes and servers, which can degrade performance in high-demand computing environments. The system includes at least two virtualization nodes, each directly connected to multiple servers and to the other virtualization node. These direct attachment connections eliminate intermediate networking layers, reducing latency and improving data transfer speeds. Each virtualization node manages virtualized resources, such as virtual machines or containers, and coordinates with the other node to balance workloads and ensure high availability. The direct connections allow for seamless failover and load distribution, enhancing system reliability and performance. The invention also includes mechanisms for dynamic resource allocation, where the virtualization nodes monitor server performance and adjust resource assignments in real-time. This ensures optimal utilization of computing resources while maintaining low-latency communication between nodes and servers. The system is particularly useful in data centers and cloud computing environments where efficiency and scalability are critical. By minimizing network hops and streamlining connections, the invention provides a more responsive and resilient virtualization infrastructure.
15. The method of claim 14 , wherein the OCS includes a first OCS and a second OCS, wherein the first OCS is coupled to a first port of the first virtualization network and to the first and second controllers, and the second OCS is coupled to a first port of the second virtualization node and to the first and second controllers.
This invention relates to network virtualization systems, specifically addressing the challenge of managing and coordinating operations across multiple virtualization nodes and controllers. The system includes a virtualization network with at least two virtualization nodes, each having multiple ports, and at least two controllers responsible for managing the virtualization nodes. The system also incorporates an Operations Control System (OCS) that interfaces with the virtualization nodes and controllers to facilitate coordinated operations. In this configuration, the OCS is divided into a first OCS and a second OCS. The first OCS is connected to a first port of the first virtualization node and communicates with both controllers, while the second OCS is connected to a first port of the second virtualization node and also communicates with both controllers. This dual OCS setup ensures redundancy and reliability in managing the virtualization network, allowing for seamless coordination between the controllers and the virtualization nodes. The system enables efficient control and monitoring of network operations, improving fault tolerance and performance in virtualized environments.
16. The method of claim 15 , wherein the first virtualization node and the second virtualization node each include an additional port to connect to at least one of the group consisting of: an additional back end storage array; and an additional server.
This invention relates to virtualization systems, specifically improving connectivity and scalability in virtualized environments. The problem addressed is the limited flexibility and expandability of virtualization nodes, which often lack sufficient ports to connect to additional storage arrays or servers, restricting system growth and resource allocation. The invention describes a method for enhancing virtualization nodes by incorporating additional ports. Each virtualization node, including a first and a second node, is equipped with an extra port designed to connect to either an additional back-end storage array or an additional server. This configuration allows for dynamic expansion of storage capacity or computational resources without requiring significant hardware modifications. The additional ports enable seamless integration of new storage arrays or servers, improving system scalability and adaptability. The solution ensures that virtualization nodes can efficiently manage increased workloads or storage demands, enhancing overall system performance and reliability. The method supports flexible resource allocation, allowing administrators to scale infrastructure as needed while maintaining operational efficiency.
17. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to: detect, via a back end path detector, a failover of a first controller of a backend storage array connected to a host computer system, which first controller is in an active state; perform a switching operation, via an optical circuit switch (OCS), based upon the detecting of the failover of the of the first controller, to connect a second controller of the backend storage array, which second controller is in a passive state, to change the second controller into an active state; and determine whether an I/O flow between the host computer system and the first controller continues after failover of the first controller has been detected, and, based on determining that the I/O flow to the first controller continues, initiating a command to the OCS to change direction of the I/O flow to the second controller.
This invention relates to high-availability storage systems, specifically addressing the problem of seamless failover in backend storage arrays to maintain uninterrupted data access for host computer systems. The system includes a backend storage array with at least two controllers—a primary (active) controller and a secondary (passive) controller—connected to a host computer system. Upon detecting a failover of the active controller, the system automatically switches the passive controller to an active state using an optical circuit switch (OCS). The OCS dynamically reroutes I/O traffic from the failed controller to the newly activated controller. Additionally, the system monitors whether I/O operations continue to flow to the failed controller after failover detection. If such traffic persists, the system issues a command to the OCS to redirect the I/O flow to the active secondary controller, ensuring continuous data access without disruption. The solution leverages real-time detection and automated switching to minimize downtime and maintain data integrity during controller failures.
Unknown
November 10, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.